[Insight-developers] FW: 2d curves extractor

[Satish]

Adding..

The human image is 512 x 512 x 1, while the phantom image is 1000 x 1000
x 1

Regards

Satish

 

From: Satish 
Sent: Thursday, January 27, 2011 1:12 PM
To: 'insight-developers at itk.org'
Subject: 2d curves extractor

 

Hi,

 

While using the 2d curves extractor, to extract curves in a XA 2d
fluoroscopic DICOM image (single slice, single frame), we find that,
when the image is of

real human with few anatomic structures, then the out image which is
supposed to contain the detected curve-points overlaid on the input
image 

actually is written well over the exact curve-structures. 

 

Whereas, when the input image is that of a phantom, then they are
written at the top-left corner

of the input image. I guess there is something to do with the different
co-ordinate space (i.e. of the outputMesh and the actual image).

It's probably after the m_InverseParametricFilter->Update();

 

But just a bit confused why it behaves differently with different input
files, although the curves are detected very well.

See attached results for human and phantom.

 

Satish Kumar B.G

 

Technical Architect

COMPUSOL

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20110127/567f34e8/attachment.htm>